There are concerns about the quality of medical care. Examples of substandard care have been found in several diseases, yet the tools to measure the quality of palliative cancer services are limited. The goal of this project is to develop a core set of valid claims-based indicators of the quality of care delivered to patients with advanced malignancies. Such indicators could provide continuous quality monitoring, and allow rapid feedback to providers so that poor quality processes of care can be identified and improved. Candidate indicators will be identified through literature review and focus groups. The feasibility of using administrative data to monitor each indicator will be assessed using SEER-Medicare linked data and local institutional data. The SEER-Medicare dataset will then be used to select measures with sufficient variability amongst Health Care Service Areas to be useful as indicators, and where possible, empiric performance benchmarks will be determined. We will determine reliability and content validity by reviewing the charts of patients in the local dataset to see if the measures obtained from billing data are accurate. Construct validity will be assessed by looking for convergent validity among the indicators. In an already ongoing study, the family members of patients who have died from cancer are being asked about the quality of care the patient received. Criterion validity will be sought by correlating the performance of each indicator as applied to these patients' billing records, with the quality perceived by their family members. Based on the results of these validation exercises, we will determine a core indicator set, and then apply them to examine the correlations of provider or patient characteristics, such as race, sociodemographics, and academic affiliation, with quality of care. This project is important because it will, for the first time, provide valid means with which to assess the quality of medical services delivered to patients with advanced cancer, and will characterize variations in quality across different types of providers. Payers and patients will benefit from these results when selecting providers, and institutions and networks will be able to use them to monitor performance.